Month: February 2019

The maths behind Product Virality and Referral Marketing

When building a consumer-facing product, virality is a key element. Virality simply converts streams of incoming customers into new streams of incoming customers, without the cost of its acquisition (or at least without part of it).

Product Virality, when done properly, allows startups to scale fast. Virality origins from the name “virus” and borrows elements like spreadability and measurable KPIs like the “K-Factor” that in Epidemiology measures the average number of people a host will contact while still infectious.

While the K-Factor is probably the most important metric when it comes to Virality, it’s still an “output” KPI that is fuelled by much more other metrics, which I’ll try to analyze in the context of this article.

Let’s start with the basics; K-Factor.

K-Factor is simply calculated by the following formula:

K -factor= i * c

i stands for the number of outgoing invitations that occurred by one existing customer and c declares the average conversion from receiving an invitation to the desired conversion (e.g. sign-up or user activation).

Although this formula is quite simple to calculate and understand, it doesn’t always tell the truth. Imagine the following edge cases:

  • The invitations happen off-line (word of mouth) and cannot be tracked
  • The product doesn’t run a traceable invitation mechanism
  • The invitations occur somewhere within the basic functionality of the product that wasn’t designed for virality (e.g. transactional emails)

There’s no perfect way, but I believe that there are two ways of calculating the K-factor more accurately to understand whether you achieve Viral Growth or not.

First one, would be to isolate the invitation activity into a traceable, identifiable mechanism. Implement tracking on invitation mechanisms (even if they don’t run incentives). Secondly, include tracking events within your product whenever the activities leak out of your echosystem. Thirdly, reverse engineer attribution and calculate potential entries to your system by word-of-mouth. Combine these three metrics into one i metric and you ‘ll have a more accurate picture

Broad Virality metrics

Active Users Sending Invites

Not all active customers behave similarly when it comes to outgoing number of invitations to prospects. There’s always a small powerful cohort that invites the majority of the referrals and wider, bigger cohort of referees that refer with ambiguous results. Tracking the size of Active Users Sending Invites is a major health-check to understand whether product changes affected or not the virality of your product.

CAC – Customer Acquisition Cost

As all traffic channels, your viral growth channel involves expenses to run as well as expenses in the form of user incentivization.

CAC = All expenses (product, marketing) involved in virality / number of new customers via referral

CLTV – Customer Life Time Value

CLTV helps us understand the quality and the value of users that are being acquired (overall or via a specific channel.

CLTV (for Referrals) = average value of a conversion x average # of conversions in a time frame x average customer lifetime

Loyal Customer Value

Customers who tend to refer more users than others are classified as loyal ones. The Loyal Customer Value should not be conflicted with the CLTV

Loyal Customer Value = Net Promoter Score/ Referrals

Customer Retention Rate (via Referrals)

Weekly, monthly or annual analysis of customers acquired via viral growth

Customer Retention Rate = ((CE-CN)/CS)) X 100

  • CE = number of users acquired via referrals at the end of a specified period
  • CN = number of users acquired via referrals during the period
  • CS = number of users acquired via referrals at the start of the period

CRR reveals churn. If CRR = 80% , Churn = 20% (and vice-versa)